Table 1.
Diagnostic Error Evaluation and Research taxonomy (total count)
| Category | ChatGPT | Human | P value |
| Access/presentation | |||
| (A) Failure/delay in presentation | 6 (1.1%) | 20 (3.7%) | 0.005 |
| (B) Failure/denied care access | 3 (0.6%) | 2 (0.4%) | 0.65 |
| History | |||
| (A) Failure/delay in eliciting critical history data | 327 (60.0%) | 29 (5.3%) | <0.001 |
| (B) Inaccurate/misinterpretation | 193 (35.4%) | 32 (5.9%) | <0.001 |
| (C) Failure in weighing | 21 (3.9%) | 33 (6.1%) | 0.09 |
| (D) Failure/delay to follow-up | 31 (5.7%) | 5 (0.9%) | <0.001 |
| Physical examination | |||
| (A) Failure/delay in eliciting critical physical examination finding | 102 (18.7%) | 18 (3.3%) | <0.001 |
| (B) Inaccurate/misinterpreted | 24 (4.4%) | 41 (7.5%) | 0.03 |
| (C) Failure in weighing | 1 (0.2%) | 21 (3.9%) | <0.001 |
| (D) Failure/delay to follow-up | 4 (0.7%) | 6 (1.1%) | 0.52 |
| Tests (laboratory/radiology) | |||
| (A) Failure/delay in ordering needed test(s) | 330 (60.6%) | 164 (30.1%) | <0.001 |
| (B) Failure/delay in performing ordered test(s) | 7 (1.3%) | 3 (0.6%) | 0.20 |
| (C) Error in test sequencing | 2 (0.4%) | 1 (0.2%) | 0.56 |
| (D) Ordering of wrong test(s) | 22 (4.0%) | 1 (0.2%) | <0.001 |
| (E) Test ordered the wrong way | 2 (0.4%) | 0 (0.0%) | 0.16 |
| (F) Sample mixup/mislabelled (eg, wrong patient/test) | 0 (0.0%) | 0 (0.0%) | N/A |
| (G) Technical errors/poor processing of specimen/test | 4 (0.7%) | 16 (2.9%) | 0.01 |
| (H) Erroneous laboratory/radiology reading of test | 72 (13.2%) | 88 (16.1%) | 0.17 |
| (I) Failed/delayed reporting of result to clinician | 40 (7.3%) | 2 (0.4%) | <0.001 |
| (J) Failed/delayed follow-up of (abnormal) test result | 189 (34.7%) | 4 (0.7%) | <0.001 |
| (K) Error in clinician interpretation of test | 257 (47.2%) | 77 (14.1%) | <0.001 |
| Assessment | |||
| (A) Failure/delay in considering the diagnosis | 510 (93.6%) | 357 (65.5%) | <0.001 |
| (B) Too little consideration/weight given to the diagnosis | 421 (77.2%) | 57 (10.5%) | <0.001 |
| (C) Too much weight on competing/coexisting diagnosis | 36 (6.6%) | 52 (9.5%) | 0.08 |
| (D) Failure/delay to recognise/weigh urgency | 50 (9.2%) | 23 (4.2%) | 0.001 |
| (E) Failure/delay to recognise/weigh complication(s) | 9 (1.7%) | 29 (5.3%) | <0.001 |
| Referral/consultation | |||
| (A) Failure/delay in ordering referral | 79 (14.5%) | 40 (7.3%) | <0.001 |
| (B) Failure/delay obtaining/scheduling ordered referral | 13 (2.4%) | 1 (0.2%) | 0.001 |
| (C) Error in diagnostic consultation performance | 8 (1.5%) | 6 (1.1%) | 0.59 |
| (D) Failure/delayed communication/follow-up of consultation | 35 (6.4%) | 6 (1.1%) | <0.001 |
| Follow-up | |||
| (A) Failure to refer patient to close/safe setting/monitoring | 17 (3.1%) | 7 (1.3%) | 0.04 |
| (B) Failure/delay in timely follow-up/rechecking of patient | 101 (18.5%) | 14 (2.6%) | <0.001 |
| Unclear | 0 (0.0%) | 3 (0.6%) | 0.08 |